DNA-Based Dating

No luck with Match.com or EHarmony? Online daters now have a new way to screen potential dates. Two companies offer genetic analysis that purports to find your perfect love match based in part on the smell of his or her sweat.

The genetic tests for both companies--ScientificMatch, based in Naples, FL, and GenePartner, based in Zurich, Switzerland--are based on the same study, performed more than a decade ago in Switzerland. Women were asked to rate the odor of men's sweaty T-shirts, and then both sexes were tested for the genotype of several genes within the major histocompatibility complex (MHC), a family of genes that is crucial for immune function. Women were more likely to prefer the sweaty smells of men whose MHC genes were most different from their own.

While it sounds bizarre, the finding does have a potential evolutionary explanation. According to an article from the Economist,

The children of couples with a wide range of MHC genes, and thus of immune responses, will be better protected from disease. . . . That could be particularly important in a collaborative, group-living species such as humanity. Moreover, comparing MHCs could be a proxy for comparing kinship, and thus help to prevent inbreeding.

Does it really work for finding a date? Probably not. Subconscious feelings about a person's smell likely make only a minor contribution to each person's attractiveness equation.

Genetic Tests Already Becoming Obsolete

Genetic testing may be even worse than flat screens when it comes to the speed at which they become out of date. Last fall, I took a genetic test for a recently identified genetic variation linked to a significantly increased risk of type 2 diabetes. With a family history of the disease and other risk factors, I was curious about my own genetic status. Fortunately, I didn't have that high-risk allele--but given my family history, I likely have other disease-linked variants.

At the Beyond Genome conference in San Francisco yesterday, I discovered that the test I took a year ago now screens for four genetic variants that boost risk for the disease. Those new additions are part of a flood of genome-wide association screens that have been reported over the past year. The rate may slow down as variants with the biggest impact on common diseases are identified.

The test I took was focused on a single disease, type 2 diabetes. Genome-wide screens, such as those offered by Navigenics, DeCode, and 23andMe, offer updates--either included in the initial fee or for an annual subscription--to their customers as additional information becomes available.

The issue led to an interesting panel discussion about who is responsible for tracking patients who do show an increased risk for disease. The problem is so new that there is no consensus yet. But Steve Murphy, founder of the personalized-medicine practice Helix Health, argued that it is the physician's job, just as it is his or her responsibility to make sure that patients with high cholesterol and other risk factors are properly monitored.

Here's an unrelated but fascinating fact that you may have heard before: a full 10 percent of fathers may not be the biological parent of their children, according to Charles Lee, director of cytogenetics at the Harvard Cancer Center. Lee's group didn't set out to examine these rates specifically--researchers do paternity testing as part of their standard genetic-testing process. In the process, they discovered that one in ten men are not the biological fathers of their children.